Statistical Modeling for Sports Prediction - an Application in the 2018 Brazilian Football Championship with the Bivariate Poisson Model
Magno Xavier de Arajo,
Anna Karla Barros da Trindade,
Francisco de Paula Santos de Araujo Junior
Issue:
Volume 11, Issue 3, June 2022
Pages:
39-46
Received:
26 April 2022
Accepted:
14 May 2022
Published:
27 June 2022
DOI:
10.11648/j.pamj.20221103.11
Downloads:
Views:
Abstract: The present work aims to approach the Holgate bivariate Poisson model in its application to soccer data, considering the importance that soccer has worldwide and as a very popular sport in Brazil. In a football match, two teams play for 90 minutes, plus extra time defined by the match referee according to the organizations' own criteria, with the objective of scoring as many goals as their opponent. It can be seen in some studies the statement that the final result of a soccer match is a bivariate random vector whose number of goals scored by each team can be represented by its parameters. In view of the above, from Holgate's bivariate Poisson model, the attack and defense capacity of each team will be estimated, based on the average of goals scored and conceded by the home team playing at home, and the visiting team playing away from home. Such a model allows the analysis of prediction of the probability of victory, draw and defeat of the teams in a certain future match, with defined parameters and a model that takes into account the goals scored and suffered by the teams until the previous round to the analysis of the model. With this, the focus is on the application of the model, to improve and perfect the ability to predict results, where the four initial rounds of the championship serve to define the initial parameters for the model. It is worth noting that the methodology presented will be applied with data from the 2018 Campeonato Brasileiro de Futebol da Série A, which is the main national championship in Brazil and involves the top twenty teams based on their rankings in the previous year's championship.
Abstract: The present work aims to approach the Holgate bivariate Poisson model in its application to soccer data, considering the importance that soccer has worldwide and as a very popular sport in Brazil. In a football match, two teams play for 90 minutes, plus extra time defined by the match referee according to the organizations' own criteria, with the o...
Show More
Generalized Form of p-Bounded Variation of Sequences of Fuzzy Real Numbers
Gyan Prasad Paudel,
Narayan Prasad Pahari,
Sanjeev Kumar
Issue:
Volume 11, Issue 3, June 2022
Pages:
47-50
Received:
8 March 2022
Accepted:
30 June 2022
Published:
12 July 2022
DOI:
10.11648/j.pamj.20221103.12
Downloads:
Views:
Abstract: Classical mathematics deals with only two conclusions: true and false. But fuzzy logic is a multiple-valued logic in which the truth values of variables might be any real number between 0 and 1. L. A. Zadeh developed the idea of fuzzy logic in 1965 to investigate the haziness and lack of concentration in information found in mathematics. The notion of the fuzzy set has been successfully applied in studying the different classes of sequence spaces. In recent years, many researchers have replaced these mathematical structures of real or complex numbers with fuzzy numbers and interval numbers and have investigated many results. This study aims to analyze the sequence space bVFp(X) for 1≤p<∞ of p- bounded variation of fuzzy real numbers and it is extended to the p- bounded variation of the difference sequence space bVFp(ΔmX) of fuzzy real numbers. The proposed study will be based on a dry lab review. It will be based on existing theories that are already proven and established. On the promise of the existing theories, we will study some new results with their different properties. To study the different properties, we will introduce a new metric on bVFp(ΔmX). Moreover, we shall explore some of the inclusion relations with respect to p and q.
Abstract: Classical mathematics deals with only two conclusions: true and false. But fuzzy logic is a multiple-valued logic in which the truth values of variables might be any real number between 0 and 1. L. A. Zadeh developed the idea of fuzzy logic in 1965 to investigate the haziness and lack of concentration in information found in mathematics. The notion...
Show More
